VIEW­POINT

Gart­ner out­lines how I&O lead­ers need to strate­gi­cally lever­age AI as a core ac­cel­er­ant to dig­i­tal busi­ness ini­tia­tives.

Network Middle East - - CONTENTS -

Gart­ner out­lines how AI can be lever­aged to ad­vance dig­i­tal busi­ness ini­tia­tives

I& O lead­ers need to strate­gi­cally lever­age AI as a core ac­cel­er­ant to dig­i­tal busi­ness ini­tia­tives.

Lead­ing or­gan­i­sa­tions ex­pect to dou­ble the num­ber of ar­ti­fi­cial in­tel­li­gence (AI) projects in place within the next year, and over 40% of them plan to ac­tu­ally de­ploy AI so­lu­tions by the end of 2020, ac­cord­ing to the Gart­ner 2020 CIO Agenda Sur­vey.

But the re­al­ity is that most or­gan­i­sa­tions strug­gle to scale the AI pi­lots into en­ter­prise wide pro­duc­tion, which lim­its the abil­ity to re­alise AI’S po­ten­tial busi­ness value.

“Launch­ing pi­lots is de­cep­tively easy but de­ploy­ing them into pro­duc­tion is no­to­ri­ously chal­leng­ing. Although the po­ten­tial for suc­cess is enor­mous, de­liv­er­ing busi­ness im­pact from AI ini­tia­tives takes much longer than an­tic­i­pated,” says Chi­rag Dekate, se­nior direc­tor an­a­lyst, Gart­ner.

“IT lead­ers re­spon­si­ble for AI are dis­cov­er­ing AI pi­lot para­dox, where launch­ing pi­lots is de­cep­tively easy but de­ploy­ing them into pro­duc­tion is no­to­ri­ously chal­leng­ing.”

IT lead­ers re­spon­si­ble for AI must nur­ture in­fra­struc­ture strate­gies that en­able the evo­lu­tion of AI pi­lots into scal­able pro­duc­tion and, im­por­tantly, value re­al­i­sa­tion. Con­sider these five pre­dic­tions in the rapid evo­lu­tion of AI tools and tech­niques and suc­cess­fully mas­ter pro­duc­tion of ar­ti­fi­cial in­tel­li­gence.

AI WILL DRIVE IN­FRA­STRUC­TURE DE­CI­SIONS

AI will re­main one of the top work­loads driv­ing in­fra­struc­ture de­ci­sions through 2023. Ac­cel­er­at­ing AI pi­lots into pro­duc­tion re­quires spe­cific in­fra­struc­ture re­sources that can grow and evolve along­side tech­nol­ogy. AI mod­els will need to be pe­ri­od­i­cally re­fined by the en­ter­prise IT team to en­sure high suc­cess rates.

This might in­clude stan­dar­d­is­ing data pipe­lines or in­te­grat­ing ma­chine learn­ing (ML) mod­els with stream­ing data sources to de­liver real-time pre­dic­tions.

MAN­AGE IN­CREAS­ING COM­PLEX­ITY OF AI TECH­NIQUES THROUGH COL­LAB­O­RA­TION

One of the top tech­nol­ogy chal­lenges in lev­er­ag­ing AI tech­niques like ML or deep neu­ral net­works (DNN) in edge and IOT (In­ter­net of Things) en­vi­ron­ments is the com­plex­ity of data and an­a­lyt­ics.

Suc­cess­fully de­ploy­ing pro­duc­tion AI in such en­vi­ron­ments will re­quire close part­ner­ship be­tween the busi­ness and IT. Proac­tively plan and pro­vide ready so­lu­tions when new busi­ness needs emerge — a con­cept Gart­ner calls in­fra­struc­ture-led dis­rup­tion.

SIM­PLE ML TECH­NIQUES SOME­TIMES MAKE THE MOST SENSE

Through 2022, more than 75% of or­gan­i­sa­tions will use DNNS for use cases that could use clas­si­cal ML tech­niques. Suc­cess­ful early AI adopters lever­aged prag­matic ML so­lu­tions to de­liver busi­ness value. These early projects used tra­di­tional sta­tis­ti­cal ma­chine learn­ing,

but as the or­gan­i­sa­tion evolved, they pur­sued more ad­vanced tech­niques with deep learn­ing to grow the im­pact of AI. Sift through the AI hype and learn the spec­trum of op­tions to ap­pro­pri­ately ad­dress busi­ness prob­lems. Opt for sim­plic­ity over pop­u­lar, but com­pli­cated, op­tions.

MAKE CLOUD SER­VICE PROVIDERS PART OF YOUR STRAT­EGY

Strate­gic use of cloud tech­nolo­gies like cog­ni­tive APIS, con­tain­ers and server­less com­put­ing can help sim­plify the com­pli­cated process of de­ploy­ing AI. By 2023, cloud-based AI will in­crease 5X from 2019, mak­ing AI one of the top cloud ser­vices. Con­tain­ers and server­less com­put­ing will en­able ML mod­els to serve as in­de­pen­dent func­tions, re­duc­ing cost and over­head.

A server­less pro­gram­ming model is par­tic­u­larly ap­peal­ing in public cloud en­vi­ron­ments be­cause of its quick scal­a­bil­ity, but IT lead­ers should iden­tify ex­ist­ing ML projects that can ben­e­fit from these new com­put­ing ca­pa­bil­i­ties.

IT lead­ers re­spon­si­ble for AI are dis­cov­er­ing AI pi­lot para­dox, where launch­ing pi­lots is de­cep­tively easy but de­ploy­ing them into pro­duc­tion is no­to­ri­ously chal­leng­ing.” Chi­rag Dekate, se­nior direc­tor an­a­lyst, Gart­ner

ADOPT AI AUG­MENTED AU­TO­MA­TION BE­YOND THE SUR­FACE LEVEL

As the amount of data that or­gan­i­sa­tions have to man­age in­creases, so too will the abun­dance of false alarms and in­ef­fec­tive prob­lem pri­ori­ti­sa­tion. It doesn’t help that IT and busi­ness units of­ten do not speak the same lan­guage when it comes to ar­ti­fi­cial in­tel­li­gence.

By em­brac­ing AI aug­mented au­to­ma­tion, IT teams can bet­ter learn the skills of ar­ti­fi­cial in­tel­li­gence and po­si­tion them­selves to have more ef­fec­tive part­ner­ships with pe­riph­eral busi­ness units.

In fact, by 2023, 40% of I& O teams will use ar­ti­fi­cial in­tel­li­gence-aug­mented au­to­ma­tion in large en­ter­prises, re­sult­ing in higher IT pro­duc­tiv­ity with greater agility and scal­a­bil­ity.

By 2023, cloud-based AI will in­crease 5X from 2019, mak­ing AI one of the top cloud ser­vices.

Newspapers in English

Newspapers from UAE

© PressReader. All rights reserved.